CN103916334B - A kind of channel estimation methods and device - Google Patents

Abstract

The embodiment of the present invention provides a kind of channel estimation methods and device, and to solve the problems, such as to carry out channel estimation according to the independent judgement in each footpath in prior art, caused conventional low signal-to-noise ratio scene lower channel estimates that performance is reduced；Methods described includes：Judge whether historical channel and present channel are close to；According to the judged result of the judgement, signal path position is obtained；According to the signal path position, current time zone channel estimation value is obtained.Method provided in an embodiment of the present invention is to be used as an entirety by channel estimation or its derivation information, further lifted noise circumstance under about whether the judgement for abrupt scene, so as to reduce the probability in activation non-signal footpath by mistake, the accuracy of channel estimation is improve.

Description

A kind of channel estimation methods and device

Technical field

The present invention relates to moving communicating field, more particularly to a kind of channel estimation methods and device.

Background technology

In mobile radio communications system, due to radio channel multi-path and time-varying transmission characteristic, channel estimation is receiver The key technology of process, which is mainly used in equilibrium and falls impact of the data to channel.Good channel estimator, needs as far as possible The support of many parameter or estimators, using these parameters or estimator as prior information, instructs current channel estimation, is answering More generally adopt with for effective diameter information, delay spread or doppler spread etc..

In various channel estimation methods, it is often more that the reception condition based on current time obtains above-mentioned key parameter Directly, efficiently and low-cost method, but when noise is larger corresponding parameter acquiring accuracy.Preferably to offset Influence of noise, is obtained more preferably channel estimation results, the acquisition of above-mentioned parameter is often carried out using historical information.

For static or gradual conventional wireless environment, above-mentioned channel estimation methods can reach better effects.However, one There is such as translational speed change or deep fade and causes life to be gone out in denier（It is phenomenon that the signal tap on certain position cuts in and out） Etc. scene when, influence whether the change of above-mentioned estimation object, so as to channel estimation produce negative impact.

To solve the impact of raw scene of going out, the first method that prior art is provided is to take taking out for each moment same position The minima of head makes decisions signal footpath based on the power of noise path as noise path.The second method that prior art is provided It is the catastrophe in the difference independent judgment root footpath of power using each footpath, is processed using respective filter coefficient.

For the first method that above-mentioned prior art is provided, as which independently carries out the judgement of each tap, and which is taken Minima, will certainly cause Noise Estimation to diminish, in low signal-to-noise ratio（Signal to Noise Ratio, SNR）Scene, can be big The big possibility for improving activation non-signal footpath by mistake.

For the second method that above-mentioned prior art is provided, in stabilization signal, either signal tap position is still Noise tap position, its difference result only power interpolation composition of Noise, will adapt to low especially for corresponding scheme SNR scenes, can cause thresholding to be difficult to rationally setting, so as to reach the purpose of respective filter coefficient reasonable employment.

In a word, the method that above-mentioned prior art is provided is the independent judgement according to each footpath, is only applicable in channel estimation Signal footpath judgement, can solve the problem that life is gone out the channel estimation problems of scene, but can cause under conventional low signal-to-noise ratio scene, believe Estimate that performance is reduced in road.

The content of the invention

To solve to carry out channel estimation and caused conventional low signal-to-noise ratio field according to the independent judgement in each footpath in prior art Scape lower channel estimates the problem that performance is reduced, and the embodiment of the present invention provides a kind of channel estimation methods and device.

The embodiment of the present invention provides a kind of channel estimation methods, including：Judge whether historical channel and present channel are close to；

According to the judged result of the judgement, signal path position is obtained；

According to the signal path position, current time zone channel estimation value is obtained；

It is described to judge whether historical channel and present channel are close to any knot using a kind of following mode or in the following manner Close：

By judging historical channel estimated value is judged with whether present channel estimated value is close to；Or

The parameter estimated using historical channel carries out equilibrium to current time data, is judged according to portfolio effect；Or

Historical data and current data are carried out related, judged according to related correlated results.

The embodiment of the present invention provides a kind of channel estimating apparatus, and described device includes：

Judge module, for judging whether historical channel and present channel are close to；

Signal footpath acquisition module, for the judged result according to the judgement, obtains signal path position；

Estimated value acquisition module, for according to the signal path position, obtaining current time zone channel estimation value；

The judge module judges historical channel using following any combinations using a kind of following mode or following manner Whether it is close to present channel.

Knowable to the embodiments of the present invention, by the judgement rather than sentencing based on independent footpath that are then based on Global Information Certainly, therefore it is more beneficial for the identification of overall signal scene changes, and is not limited only to the use of the judgement of effective diameter, also includes Other channel estimations need the parameter that accurately could be obtained by cycle statistics, such as Doppler's estimated value.The embodiment of the present invention is carried For method be that an entirety is used as by channel estimation or its derivation information, further lifted noise circumstance under about whether For the judgement of abrupt scene, so as to reduce the probability in activation non-signal footpath by mistake, the accuracy of channel estimation is improve.

Description of the drawings

In order to be illustrated more clearly that the technical scheme of the embodiment of the present invention, below will be in prior art or embodiment description The required accompanying drawing for using is briefly described, it should be apparent that, drawings in the following description are only some realities of the present invention Example is applied, for those skilled in the art, other accompanying drawings can also be obtained such as these accompanying drawings.

Fig. 1 is a kind of channel estimation methods schematic flow sheet provided in an embodiment of the present invention；

Fig. 2 is a kind of channel estimating apparatus structural representation provided in an embodiment of the present invention；

Fig. 3 is the channel estimating apparatus structural representation that another embodiment of the present invention is provided；

Fig. 4-a are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Fig. 4-b are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Fig. 5-a are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Fig. 5-b are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Fig. 6-a are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Fig. 6-b are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Fig. 7 is the channel estimating apparatus structural representation that another embodiment of the present invention is provided；

Fig. 8-a are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Fig. 8-b are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Fig. 9-a are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Fig. 9-b are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Figure 10 is the channel estimating apparatus structural representation that another embodiment of the present invention is provided；

Figure 11-a are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Figure 11-b are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Figure 12-a are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Figure 12-b are the channel estimating apparatus structural representations that another embodiment of the present invention is provided；

Figure 12-c are the channel estimating apparatus structural representations that another embodiment of the present invention is provided.

Specific embodiment

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than the embodiment of whole.It is based on Embodiment in the present invention, the every other embodiment obtained by those skilled in the art, belongs to the scope of protection of the invention.

Accompanying drawing 1 is referred to, is a kind of channel estimation methods schematic flow sheet provided in an embodiment of the present invention, it is main to include step Rapid S101, step S102, step S103 and step S104, specifically：

S101, judges whether historical channel and present channel are close to.

In embodiments of the present invention, can be by judging historical channel estimated value is sentenced with whether present channel estimated value is close to Whether disconnected historical channel and present channel are close to, or the parameter estimated using historical channel is carried out to current time data According to portfolio effect, weighing apparatus, judges whether historical channel is close to present channel, or historical data is carried out phase with current data Close, according to related correlated results, judge whether historical channel is close to present channel, or appointing in above-mentioned three kinds of modes Several pictures of anticipating are combined, and compound mode of the present invention is not limited.

First, illustrate that the real-time tap power of time domain channel estimates value sequence P（k）, the real-time tap power of time domain channel Estimate value sequence P（k）When representing current time tap power value sequence subsequent time, the tap power values at a upper moment are averagely arrived In each tap average channel estimation power value sequence P_mean (k) of history, the moment tap power value sequence is recorded again.At this In inventive embodiments, the performance number of each tap can be averaging and value sequence is estimated with the real-time tap power for obtaining time domain channel P（k）.A frame data can be received first, obtain the OFDM comprising reference signal（Orthogonal Frequency Division Multiplexing, OFDM）The data of symbol.In embodiments of the present invention, reference signal can be Long Term Evolution （Long Term Evolution, LTE）Reference signal in system frame structure, a frame of LTE have 14 symbols, wherein referring to Signal is located on the 0th, 4,7,11 symbol.The number of reference signal is relevant with the type of bandwidth and reference signal, as a width of 20M of band When, cell-specific reference has 200.

The performance number of each tap is averaging to obtain the real-time tap power estimation value sequence P of time domain channel（k）Can be with Completed based on the least-squares estimation of 4 reference signals, specifically, for each symbol containing reference signal, extracted with reference to letter Number, the channel estimation of the frequency domain of least square being completed, after being transformed into time domain, 4 groups of time domain channels is obtained and is estimated, calculating is respectively taken out The performance number of head, and the para-position performance number for carrying out four symbols by tap position is average, the final real-time tap for obtaining time domain channel Power estimation value sequence P（k）, herein, k=1,2 ..., N, N are channel estimation length, when value is with the maximum possible of setting is expected Prolong extension it is relevant, in embodiments of the present invention, N can value be 200.

As the present invention by judging historical channel estimated value judges historical channel with whether present channel estimated value is close to The one embodiment whether being close to present channel, can be each tap average channel estimation power value sequence P_ of usage history mean（k）Value sequence P is estimated with the real-time tap power of the time domain channel of aforementioned acquisition（k）Normalization related operation is carried out, is obtained Obtain normalization factor alpha.If normalization coefficient α is more than or equal to the first threshold value, judge that historical channel is close to present channel, Otherwise, judge that historical channel is kept off with present channel.In embodiments of the present invention, normalization factor alpha can be by following expression Formula is tried to achieve：

In expression formula, the meaning of N is identical with previous embodiment, i.e. N is channel estimation length.

It should be noted that above-mentioned first threshold value can be obtained by emulation mode, it is also possible to determine based on experience value. In embodiments of the present invention, if the first threshold value is represented using th1, it is 0 in the scope of th1<Th1≤1, for example, th1 can be with It is taken as 0.75.

As the present invention by judging historical channel estimated value judges historical channel with whether present channel estimated value is close to Another embodiment whether being close to present channel, can estimate value sequence P according to real-time tap power（k）History it is mean square Root delay spread value T_rms_H estimates value sequence P with real-time tap power（k）Root-mean-square delay spread value T_rms difference with Second threshold value（Represented using th2）Relation and real-time tap power estimate value sequence P（k）History delay estimation maximum Value T_max_H estimates value sequence P with real-time tap power（k）Delay estimation maximum value T_max difference and the 3rd threshold value（Make Represented with th3）Relation judge, specifically, if the absolute value of the difference of T_rms_H and T_rms be more than the second threshold value and The absolute value of the difference of T_max_H and T_max is more than the 3rd threshold value, then judge that historical channel is kept off with present channel, no Then, judge that historical channel is close to present channel.In embodiments of the present invention, T_rms_H, T_max_H, T_rms and T_max are equal Can be by real-time tap power be estimated value sequence P（k）In non-signal footpath zero setting obtain, except that, T_rms_H and T_max_H is according to tap power estimates value sequence P in real time before（k）In non-signal footpath zero setting obtain, belong to history value, its Default value can be 0, and T_rms and T_max are to estimate value sequence P according to real-time tap power（k）In non-signal footpath put Zero obtains.Similar with the first threshold value, in embodiments of the present invention, the second threshold value and the 3rd threshold value can also be by emulation Mode is obtained or is determined based on experience value.Specifically, the span of the second threshold value th2 can to expected given maximum 0 Can be worth, for example, according to 3GPP standards, it can be 5000ns that the second threshold value th2 is expected given maximum value possible（Nanosecond）, and 3rd threshold value th3 span is 0<th3≤th2.Used as one embodiment of the invention, the second threshold value th2 can obtain and be 325.5208ns（Nanosecond）, the 3rd threshold value th3 can be taken as 130.2083ns（Nanosecond）.

As the present invention by judging historical channel estimated value judges historical channel with whether present channel estimated value is close to Another embodiment whether being close to present channel, can be the channel estimation power P_role to different time points（k）Enter Row para-position is averagely with current strong footpath sequence pos of acquisition（s）If, the virtual value of strong footpath sequence pos_history of history（s）With it is current Strong footpath sequence pos（s）Virtual value differ, then judge that historical channel is kept off with present channel, otherwise, judge historical channel It is close to the present channel.Specifically, a time slot data are received, then, training sequence is extracted（That is midamble）Data, According to local ident, Steiner channel estimations are completed, this subscriber channel is extracted according to this user profile and estimates Che（k）, and try to achieve The channel estimation power P_role of different time points（k）, herein, k=1,2 ..., N, in the present embodiment, N can be taken as 16；On It is this channel impulse response acquisition process to state.To P_role（k）Stored according to sliding window storing mode, the present embodiment In, sliding window a length of 4 is taken, channel estimation length is 16.According to the storage element at current time, the channel of different time points is estimated Meter power P _ role（k）Carry out para-position and averagely obtain real-time tap power estimation value sequence P of short cycle at nearest moment（k）, obtain S most strong footpath, and its serial number is recorded, it is designated as pos（s）, pos（s）Represent current strong footpath sequence, s=1,2 ..., S, S Value is 2～6 integer（The number of the effective diameter that expresses possibility）, and judge pos（s）Corresponding P on position（k）Whether it is more than and makes an uproar Th times of acoustical power, th are the number more than 0, by pos if being not more than（s）It is set to invalid value.The present embodiment S selects to select for 2, th 0.3 is selected as, noise power obtains 16 times of acquisitions of the summation power for taking last 4 footpaths of sequence；Above-mentioned is the extraction of effective diameter Process.Further, judge strong footpath sequence pos_history of history（s）With current strong footpath sequence pos（s）Virtual value whether It is identical, wherein, virtual value different definition is for any one pos（s）In virtual value, in pos_history（s）In look for not To the element of identical Effective Numerical, or for any one pos（s）Virtual value in pos_history（s）In can not find it is identical The element of Effective Numerical.If strong footpath sequence pos_history of history（s）With current strong footpath sequence pos（s）Virtual value it is different, Then judge that present channel is kept off with historical channel, otherwise, judge that present channel is close to historical channel.

In fact, judging historical channel above by judging whether historical channel estimated value is close with present channel estimated value The embodiment whether being close to present channel, using the power based on the cumulative channel estimation of periodicity, is extracted strong path position, is carried out Current channel information and the similarity judgement of historical channel information, for control and the judgement of conduct effective diameter, obtain accordingly Signal path position, completes complete channel estimation by channel estimation post processing.

Historical channel and currently is judged as by judging whether historical channel estimated value and present channel estimated value are close to The one embodiment whether channel is close to, can carry out equilibrium according to local training sequence and to the data in training sequence domain before Data afterwards, obtain Signal to Interference plus Noise Ratio, if the Signal to Interference plus Noise Ratio be more than or equal to the 4th threshold value, judge the historical channel with The present channel is close to.Specifically, a time slot data are received first, then, extract midamble data, according to local ID, completes Steiner channel estimations, extracts this subscriber channel according to this user profile and estimates Che (k), and tries to achieve different time The channel estimation power P_role of point（k）, herein, k=1,2 ..., N, in the present embodiment, N can be taken as 16；Above-mentioned is this Channel impulse response acquisition process.If now previous moment channel estimation is present, historical channel impulse response is borrowed to this The data of data field part or the data of training sequence domain part carry out equilibrium；If original state, i.e., without previous moment channel Estimate, then, after routinely scheme completes channel estimation, preserve historical channel impulse response, this is based on the equal of historical channel estimation Weighing apparatus is obtained.The present embodiment using it is previous when channel estimation the data of training sequence domain part are carried out with channel equalization, obtain balanced Data afterwards, further according to the data of local training sequence, obtain Signal to Interference plus Noise Ratio SINR.If SINR is more than or equal to the 4th threshold value （Represented using th4）, then judge that present channel is close to historical channel；If SINR is less than the 4th threshold value, by this letter The power tap of channel shock response obtains signal to noise ratio（Signal Noise Ratio, SNR）Estimate；If SNR is more than the 6th thresholding Value（Represented using th6）, then judge that present channel is kept off with historical channel；If SNR is less than or equal to the 6th threshold value（Use Th6 is represented）, then judge that present channel is close to historical channel；In the present embodiment, th4 can be with value as -8db ~ 2db, example Such as, th4 can be with value as -4db, and th6 can be with value as -10db ~ -4db, and for example, th6 is taken as 7db.The present embodiment be with As a example by the channel estimation of TD-SCDMA system, using the hypothesis condition based on channel similarity, by historical channel estimation balancing The Signal to Interference plus Noise Ratio of this secondary data（Signal Interference Noise Ratio, SINR）, obtain current channel information and go through The similarity judgement of history channel information, implements effective diameter judgement for instructing and controlling（It is mainly used in the judgement for emptying）, obtain Whether corresponding signal path position, mainly counted superposition channel estimation by SINR, after channel estimation herein again Process completes complete channel estimation.

It is with the channel estimation of TD-SCDMA system, related to this reception training data by historical reception training data Property, using the computing mode of low-cost, the similarity for obtaining current channel information with historical channel information is adjudicated, for instructing and Effective diameter judgement is implemented in control, obtains corresponding signal path position, then completes complete channel estimation by channel estimation post processing As a example by.In TD-SCDMA frame structures, TD-SCMDA is 7 regular time slots and 3 special time slots, in 3 special time slots DwPTS, comprising fixed synchronous training code word（Represented using SYNC_DL）, and which will not change with user profile.As Historical data and current data are carried out related, judge whether historical channel is close to present channel according to related correlated results One embodiment, Ke Yishi：By synchronous training code word SYNC_DL_history of history（k）Virtual value is instructed with current sync Practice code word SYNC_DL（k）Related operation is normalized, related operation factor beta is obtained, related operation factor beta is

Herein, SYNC_DL (k）It is to take out from the data of the DwPTS for receiving, k=1,2 ..., K, SYNC_DL（k）Original length Spend for 64, be to retain multipath extension redundancy, K spans are 64～80, and for example, N can be taken as 64 in the present embodiment.Need It is bright, if SYNC_DL（k）Without original state, i.e., without previous moment history value, then by this SYNC_DL（k）Storage, makees For follow-up SYNC_DL_history（k）.If related operation factor beta is more than or equal to the 5th threshold value, present channel is judged It is close to historical channel；Otherwise, judge that present channel is kept off with historical channel.The span of the 5th threshold value 0 ~ 1 it Between, for example, 0.7 is can be taken as in the present embodiment.If related operation factor beta is less than the 5th threshold value, by this SYNC_DL （k）Channel estimation power tap obtain SNR estimated values.When SNR estimated values are less than or equal to the 7th threshold value, then judge current Channel is close to historical channel.The span of the 7th threshold value for example, in the present embodiment may be used between -10db ~ -4db It is taken as -7db.

S102, according to the judged result for judging whether historical channel and present channel are close to, obtains signal path position.

Judging historical channel and ought above by judging whether historical channel estimated value and present channel estimated value are close to In the embodiment whether front channel is close to, as according to the judged result for judging whether historical channel and present channel are close to, obtain One embodiment of the number of winning the confidence path position, Ke Yishi：If judging, historical channel is close to present channel, by each tap of history Average channel estimation power value sequence P_mean（k）Value sequence P is estimated with real-time tap power（k）Carry out para-position summation and obtain new Power value sequence P ' _ mean（k）, then by new power value sequence P ' _ mean（k）Sequence of the intermediate value more than P ' max_mean/3 Position is taken as signal path position, and herein, P ' max_mean are new power value sequence P ' _ mean（k）Middle maximum, and P ' Max_mean/3 represents the 1/3 of maximum P ' max_mean；If judging, historical channel is kept off with present channel, is taken out in real time Head power estimation value sequence P（k）Intermediate value is taken as signal path position more than the sequence position of Pmax/3, wherein, Pmax is to work as Front time domain channel power rough estimate value sequence P（k）Middle maximum, Pmax/3 for maximum Pmax 1/3.

It should be noted that in the embodiments of the present invention, additionally providing to P_mean（k）, T_rms_H and T_max_ The update mechanism of the historical informations such as H.For example, for each tap average channel estimation power value sequence P_mean of history（k）, When present channel is close with historical channel, storage element array P_all is estimated to historical channel（T, k）It is updated, herein, t= 1,2 ..., T, k=1,2 ..., N.T is natural number, represents that sliding window is long, and in inventive embodiments, T can be taken as 4.History is believed Estimate storage element array P_all in road（T, k）The principle being updated is：When t negated 1 is worth, k=1,2 ..., N, P_all（T, k） The temporal history storage channel information of t-1 before expression；If the P_all of t-1 before（T, k）Lack, then P_all（T, k）Default value be 0.When t takes 1, P_all（T, k）For the storage information being currently put into, its P_all（T, k）With one a pair of P (k) Should.

When present channel is kept off with historical channel, for P_all（T, k）Middle t takes 1 element, the storage letter being put into Breath be all zero, t for other values when, P_all（T, k）The temporal history storage channel information of t-1 before expression；If it The P_all of front t-1（T, k）Lack, then P_all（T, k）Default value be 0.

According to P_all（T, k）, all t corresponding to any k are carried out averagely, to obtain the average channel of each tap of history Power value sequence P_mean of estimation（k）, here it is power value sequence P_mean of the average channel estimation of each tap of history（k） Renewal process.

According to P_all（T, k）, all t corresponding to any k are carried out averagely, to obtain the average channel of each tap of history Power value sequence P_mean of estimation（k）, according to P_mean（k）Obtain history delay estimation maximum value T_max_H and history is equal Side's value delay spread value T_rms_H, here it is the renewal process of T_max_H and T_rms_H.

For judging historical channel and currently by judging whether historical channel estimated value and present channel estimated value are close to The embodiment whether channel is close to, as according to the judged result for judging whether historical channel and present channel are close to, obtains letter One embodiment of number path position, can be the channel estimation power P_role that different time points are obtained according to user profile（k）； If historical channel is close to present channel, by each tap average channel estimation power value sequence P_mean of history（k）With institute State the channel estimation power P_role of different time points（k）Carry out para-position summation and obtain new power value sequence P ' _ mean（k）, By new power value sequence P ' _ mean（k）Intermediate value is taken as the signal footpath more than the sequence position of the Pmax/3 Position, the P ' max_mean are new power value sequence P ' _ mean（k）Middle maximum, the P ' max_mean/3 are The 1/3 of the maximum P ' max_mean；If historical channel is kept off with present channel, short at the nearest moment cycle is taken out in real time Head power estimation value sequence P（k）Intermediate value is taken as the signal path position more than the sequence position of the Pmax/2, described Pmax is that real-time tap power of short cycle at the nearest moment estimates value sequence P（k）Middle maximum, the Pmax/2 for it is described most The 1/2 of big value Pmax, meanwhile, by the P of un-activated position（k）Zero setting, is easy to use during follow-up storage.

It is similar with previous embodiment, in the present embodiment, the renewal to historical information is also provided, for example, for history Channel estimation storage element array P_all（T, k）If present channel is close to historical channel, historical channel is estimated to store Cell array P_all（T, k）Be updated, herein, t=1,2 ..., T, k=1,2 ..., N, T be natural number, represent sliding window It is long, 16 can be taken as in the present embodiment T.Storage element array P_all is estimated to historical channel（T, k）The principle of renewal is：T takes During non-1 value, k=1,2 ..., N, P_all（T, k）The temporal history storage channel information of t-1 before expression；If t- before 1 P_all（T, k）Lack, then P_all（T, k）Default value be 0.When t takes 1, P_all（T, k）For the storage letter being currently put into Breath, its P_all（T, k）Correspond with P (k).When present channel is kept off with historical channel, for P_all（T, k）Middle t Take 1 element, the storage information being put into be all zero, t for other values when, P_all（T, k）Before expression, t-1's is temporal History storage channel information；If the P_all of t-1 before（T, k）Lack, then P_all（T, k）Default value be 0.

For judging historical channel and currently by judging whether historical channel estimated value and present channel estimated value are close to The embodiment whether channel is close to, according to its judged result, obtains signal path position, Ke Yishi：Channel to different time points Estimate power P _ role（k）Carry out para-position averagely value sequence P is estimated to obtain real-time tap power of short cycle at nearest moment（k）； If historical channel is close to present channel, by each tap average channel estimation power value sequence P_mean of history（k）With most Closely real-time tap power of short cycle at moment estimates value sequence P（k）Carry out para-position summation and obtain new power value sequence P ' _ mean （k）；By new power value sequence P ' _ mean（k）Intermediate value is taken as signal path position more than the sequence position of Pmax/3, its In, P ' max_mean are new power value sequence P ' _ mean（k）Middle maximum, P ' max_mean/3 are maximum P ' max_ The 1/3 of mean；If historical channel is kept off with present channel, real-time in short at nearest moment cycle tap power is estimated into value sequence P（k）Intermediate value is taken as signal path position more than the sequence position of Pmax/2, and herein, Pmax is that the short cycle at nearest moment is real-time Tap power estimates value sequence P（k）Middle maximum, Pmax/2 for maximum Pmax 1/2, meanwhile, by the P of un-activated position （k）Zero setting, is easy to use during follow-up storage.

In the above-described embodiments, the renewal to historical information is also provided, for example, for historical channel estimates storage element Array P_all（T, k）If present channel is close to historical channel, storage element array P_all is estimated to historical channel（T, k）Be updated, herein, t=1,2 ..., T, k=1,2 ..., N, T be natural number, represent that sliding window is long, can be with the present embodiment T It is taken as 16.Storage element array P_all is estimated to historical channel（T, k）The principle of renewal is：When t negated 1 is worth, k=1,2 ..., N, P_all（T, k）The temporal history storage channel information of t-1 before expression；If the P_all of t-1 before（T, k）Lack Lose, then P_all（T, k）Default value be 0.When t takes 1, P_all（T, k）For the storage information being currently put into, its P_all（T, k） Correspond with P (k).When present channel is kept off with historical channel, for P_all（T, k）Middle t takes 1 element, is put into Storage information be all zero, t for other values when, P_all（T, k）The temporal history storage channel letter of t-1 before expression Breath；If the P_all of t-1 before（T, k）Lack, then P_all（T, k）Default value be 0.

For another example, for strong footpath sequence pos_history of history（s）The renewal of position, can be according to P_all（T, k）To appointing All t corresponding to meaning k carry out averagely obtaining power value sequence P_mean of the average channel estimation of each tap of history（k）, obtain S most strong footpath is obtained, and records its serial number, be designated as strong footpath sequence pos_history (s) of history, the value of s=1..S, S is 2 ～6 integer（The number of the effective diameter that expresses possibility）, and judge that whether corresponding P (k) is more than noise power on pos (s) positions Th ' times, th ' herein is the number more than 0, and pos_history (s) is set to invalid value if no more than 0.The present embodiment In, S is taken as 2, th ' 0.25 is taken as, noise power obtains 2 times of acquisitions of the summation power in 4 footpaths after taking.

For it is above-mentioned historical data and current data are carried out it is related, according to related correlated results judge historical channel and The embodiment whether present channel is close to, as according to the judged result for judging whether historical channel and present channel are close to, obtains One embodiment of the number of winning the confidence path position, Ke Yishi：If judging, present channel is close to historical channel, according to the training of superposition The activation footpath of the Steiner channel estimation powers of sequence obtains the signal path position of the channel estimation of the training sequence at lower moment； If judging, present channel is kept off with historical channel, according to the Steiner channel estimation powers of the training sequence at the moment Activation footpath obtains the signal path position of the channel estimation of the training sequence at lower moment.Specifically, midamble data are received, according to Local ident, completes Steiner channel estimations, extracts this subscriber channel according to this user profile and estimates Che（k）, and try to achieve difference The channel estimation power P_role of time point（k）, herein, k=1,2 ..., N, in the present embodiment, N can be taken as 16；To difference The channel estimation power P_role of time point（k）It is averagely rough to obtain short cycle time domain channel power of nearest moment to carry out para-position Estimate value sequence P（k）；If historical channel is close to present channel, by each tap average channel estimation performance number sequence of history Row P_mean（k）Cycle time domain channel power rough estimate value sequence P short with the nearest moment（k）Carry out para-position summation and obtain new Power value sequence P ' _ mean（k）；By new power value sequence P ' _ mean（k）Intermediate value is taken more than the sequence position of Pmax/3 For signal path position, wherein, P ' max_mean are new power value sequence P ' _ mean（k）Middle maximum, P ' max_mean/3 are The 1/3 of maximum P ' max_mean.If judging, present channel is kept off with the historical channel, according to the training sequence at the moment The activation footpath of the Steiner channel estimation powers of row obtains the signal path position of the channel estimation of the training sequence at lower moment, tool Body ground, receives midamble data first, according to local ident, completes Steiner channel estimations, extract this according to this user profile Subscriber channel estimates Che（k）, and try to achieve the channel estimation power P_role of different time points（k）, herein, k=1,2 ..., N, In the present embodiment, N can be taken as 16；By short at nearest moment cycle time domain channel power rough estimate value sequence P（k）Intermediate value is more than The sequence position of Pmax/2 is taken as signal path position, and herein, Pmax is that short cycle time domain channel power of nearest moment is rough Estimate value sequence P（k）Middle maximum, Pmax/2 for maximum Pmax 1/2, meanwhile, by the P of un-activated position（k）Zero setting, just Use when follow-up storage.

It is similar with previous embodiment, in the present embodiment, the renewal to historical information is also provided, for example, for synchronization Training code word SYNC_DL（k）Renewal, when present channel and historical channel are close, the required historical information of channel estimation Storage it is constant.The present embodiment is the Steiner at each moment of the finite observation time window of the channel estimation based on midamble Channel estimation power.If present channel is kept off with historical channel, the Steiner channel estimations of the midamble of storage are emptied Power.Storage current sync training code word SYNC_DL（k）, as synchronous training code word SYNC_DL_history of history（k）.

S103, according to signal path position, obtains current time zone channel estimation value.

Judging historical channel and ought above by judging whether historical channel estimated value and present channel estimated value are close to In the embodiment whether front channel is close to, as the present invention according to the signal path position, current time zone channel estimation value is obtained One embodiment, can be directly the corresponding power value sequence of the number of winning the confidence path position as current time zone channel estimation value, Maximum delay expanding value and root-mean-square delay spread value can be calculated according to the signal path position, further according to it is described maximum when Prolong expanding value and root-mean-square delay spread value, obtain current time zone channel estimation value；Or, can be with according to judging historical channel Doppler's estimated value is obtained with the judged result whether present channel is close to, and according to the signal path position, when calculating maximum Prolong expanding value and root-mean-square delay spread value, further according to Doppler's estimated value, maximum delay expanding value and root-mean-square time delay Expanding value, obtains current time zone channel estimation value.Wherein, as sentencing according to judging whether historical channel and present channel were close to Disconnected result obtains one embodiment of Doppler's estimated value, if can judge that historical channel is close to present channel, be based on The Doppler of history multiframe sample estimates to obtain Doppler's estimated value, if judging, historical channel is kept off with present channel, takes Acquiescence Doppler estimated value is Doppler's estimated value.In embodiments of the present invention, give tacit consent to Doppler's estimated value and can be taken as 0.

In the embodiments of the present invention, a kind of mode is the power information and history of the channel estimation using current slot The power information of channel estimation carry out normalization related operation and carry out the judgement of channel relevancy, and the feelings by the dependency Condition, controls the acquisition of effective diameter judgement, thus further obtains delay spread and estimates to estimate with Doppler.Another way is to adopt With this statistical variable estimated based on overall channel of delay spread, the similarity judgement of current information and historical information is carried out, The acquisition of control effective diameter judgement, thus further obtains delay spread and estimates to estimate with Doppler, and complete according to above- mentioned information Into channel estimation.

For judging historical channel and currently by judging whether historical channel estimated value and present channel estimated value are close to The embodiment whether channel is close to, which obtains current time zone channel estimation value, Ke Yishi according to signal path position：According to signal Path position, the subscriber channel to being extracted according to user profile estimate power Che（k）Middle non-signal path position value zero setting, to obtain Current time zone channel estimation value.It is similar with previous embodiment, in the present embodiment, also provide the renewal to historical information, example Such as, for historical channel estimates storage element array P_all（T, k）If present channel is close to historical channel, history is believed Estimate storage element array P_all in road（T, k）Be updated, herein, t=1,2 ..., T, k=1,2 ..., N, T be natural number, table Show that sliding window is long, 16 can be taken as in the present embodiment T.Storage element array P_all is estimated to historical channel（T, k）The original of renewal It is then：When t negated 1 is worth, k=1,2 ..., N, P_all（T, k）The temporal history storage channel information of t-1 before expression； If the P_all of t-1 before（T, k）Lack, then P_all（T, k）Default value be 0.When t takes 1, P_all（T, k）Currently to put The storage information for entering, its P_all（T, k）Correspond with P (k).When present channel is kept off with historical channel, for P_ all（T, k）Middle t takes 1 element, the storage information being put into be all zero, t for other values when, P_all（T, k）T- before expression 1 temporal history storage channel information；If the P_all of t-1 before（T, k）Lack, then P_all（T, k）Default value be 0。

It is related for historical data and current data are carried out, according to related correlated results judge historical channel with it is current The embodiment whether channel is close to, which obtains current time zone channel estimation value, Ke Yishi according to signal path position：According to described Signal path position, the value of the activation path position of Steiner channel estimations is retained, and other values zero setting is obtaining the current time zone Channel estimation value.

It is similar with previous embodiment, in the present embodiment, the renewal to historical information is also provided, for example, for synchronization Training code word SYNC_DL（k）Renewal, when present channel and historical channel are close, the required historical information of channel estimation Storage it is constant.The present embodiment is the Steiner at each moment of the finite observation time window of the channel estimation based on midamble Channel estimation power.If present channel is kept off with historical channel, the Steiner channel estimations of the midamble of storage are emptied Power.Storage current sync training code word SYNC_DL（k）, as synchronous training code word SYNC_DL_history of history（k）.

Knowable to the channel estimation methods that the embodiments of the present invention are provided, by the judgement for being then based on Global Information not It is the judgement based on independent footpath, therefore is more beneficial for the identification of overall signal scene changes, and is not limited only to effective diameter The use of judgement, also needs the parameter that accurately could be obtained by cycle statistics, such as Doppler to estimate comprising other channel estimations Value.Method provided in an embodiment of the present invention is to be used as an entirety by channel estimation or its derivation information, is further lifted and is made an uproar Under acoustic environment about whether the judgement for abrupt scene, so as to reduce the probability in activation non-signal footpath by mistake, improve channel The accuracy of estimation, at the same time, while SYNC_DL possesses the advantage of low-cost.

Accompanying drawing 2 is referred to, is channel estimating apparatus structural representation provided in an embodiment of the present invention.For convenience of description, Illustrate only the part related to the embodiment of the present invention.The channel estimating apparatus of 2 example of accompanying drawing include judge module 201, letter Number footpath acquisition module 202 and estimated value acquisition module 203, wherein：

Judge module 201, for judging whether historical channel and present channel are close to；

Signal footpath acquisition module 202, for the judged result according to the judge module 201, obtains signal path position；

Estimated value acquisition module 203, for according to the signal path position, obtaining current time zone channel estimation value；

Wherein, judge module 201 adopts a kind of following mode or any combination of following manner to judge historical channel and work as Whether front channel is close to：Judge historical channel and ought by judging whether historical channel estimated value and present channel estimated value are close to Whether front channel is close to；Or the parameter estimated using historical channel carries out equilibrium to current time data, according to portfolio effect Judge whether historical channel is close to present channel；Or historical data and current data are carried out it is related, according to related phase Close result and judge whether historical channel is close to present channel.

It should be noted that in the embodiment of above channel estimating apparatus, the division of each functional module is only to illustrate It is bright, can as needed in practical application, such as the convenient consideration of the realization of the configuration requirement or software of corresponding hardware, and incite somebody to action Above-mentioned functions distribution is completed by different functional modules, will the internal structure of the channel estimating apparatus be divided into different work( Energy module, to complete all or part of function described above.And, the corresponding work(in practical application, in the present embodiment Energy module can be realized by corresponding hardware, it is also possible to perform corresponding software by corresponding hardware and complete, for example, aforesaid Signal footpath acquisition module, can be aforementioned according to the judge module with performing（Or diagnosis apparatuss）Judged result, obtain letter The hardware of number path position, for example, signal footpath getter, or be able to carry out corresponding computer program so as to complete aforementioned work( The general processor or other hardware devices of energy；For another example aforesaid estimated value acquisition module, can be with aforementioned of execution According to the signal path position, the hardware of current time zone channel estimation value function, such as estimated value getter, or energy are obtained Corresponding computer program is performed enough so as to complete the general processor or other hardware devices of aforementioned function（This specification is provided Each embodiment can all apply foregoing description principle）.

The judge module 201 of 2 example of accompanying drawing can include that the first judging unit 301, the second judging unit 302 or the 3rd are sentenced Disconnected unit 303, the channel estimating apparatus that another embodiment of the present invention is provided as shown in Figure 3, wherein：

First judging unit 301, if being more than or equal to the first threshold value for normalization coefficient α, judges the history Channel is close to present channel, otherwise, judges that the historical channel is kept off with present channel, and the normalization coefficient α is equal to

The N is channel estimation length；

Second judging unit 302, if for T_rms_H and the difference of T_rms absolute value be more than the second threshold value and The absolute value of the difference of T_max_H and T_max is more than the 3rd threshold value, then judge that the historical channel is not connect with present channel Closely, otherwise, judge that the historical channel is close to present channel, the T_rms_H is that real-time tap power is estimated value sequence P （k）The history root-mean-square delay spread value tried to achieve after the zero setting of middle non-signal footpath, the T_rms are by real-time tap power estimated value Sequence P（k）The root-mean-square delay spread value tried to achieve after the zero setting of middle non-signal footpath, the T_max_H is to estimate real-time tap power Evaluation sequence P（k）The history delay estimation maximum value tried to achieve after the zero setting of middle non-signal footpath, the T_max is to be taken out described in real time Head power estimation value sequence P（k）The delay estimation maximum value tried to achieve after the zero setting of middle non-signal footpath；

3rd judging unit 303, carries out para-position for the channel estimation power to different time points current averagely to obtain Strong footpath sequence pos（s）If, the virtual value of strong footpath sequence pos_history of history（s）With current strong footpath sequence pos（s）'s Virtual value is differed, then judge that the historical channel is kept off with the present channel, otherwise, judges the historical channel and institute State present channel to be close to.

The signal footpath acquisition module 202 of 3 example of accompanying drawing 2 or accompanying drawing can include the first para-position sum unit 401, such as accompanying drawing The channel estimating apparatus that 4-a or another embodiment of the present invention shown in accompanying drawing 4-b are provided.If the first para-position sum unit 401 is used for The judge module 201 judges that historical channel is close to present channel, then by each tap average channel estimation performance number of history Sequence P_mean（k）Value sequence P is estimated with real-time tap power（k）Carry out para-position summation and obtain new power value sequence P ' _ mean（k）If the historical channel is kept off with the present channel, real-time tap power is estimated into value sequence P（k）In Value is taken as the signal path position more than the sequence position of the Pmax/3, and the Pmax is real-time tap power estimated value Sequence P（k）Middle maximum, the Pmax/3 are the 1/3 of the maximum Pmax, by new power value sequence P ' _ mean （k）Intermediate value is taken as the signal path position more than the sequence position of P ' max_mean/3, and the P ' max_mean are described New power value sequence P ' _ mean（k）Middle maximum, the P ' max_mean/3 for the maximum P ' max_mean 1/3.

The estimated value acquisition module 203 of 3 example of accompanying drawing 2 or accompanying drawing can directly take the corresponding power of the signal path position Value sequence is used as the current time zone channel estimation value.

The estimated value acquisition module 203 of 3 example of accompanying drawing 2 or accompanying drawing can also be obtained including the first computing unit 501 and first Unit 502 is taken, the channel estimating apparatus that such as accompanying drawing 5-a or another embodiment of the present invention shown in accompanying drawing 5-b are provided, wherein：

First computing unit 501, for according to the signal path position, calculating maximum delay expanding value and root-mean-square time delay Expanding value；

First acquisition unit 502, for according to the maximum delay expanding value and root-mean-square delay spread value, obtaining current Time domain channel estimated value.

The estimated value acquisition module 203 of 3 example of accompanying drawing 2 or accompanying drawing can also be calculated including second acquisition unit 601, second The channel estimation that unit 602 and the 3rd acquiring unit 603, such as accompanying drawing 6-a or another embodiment of the present invention shown in accompanying drawing 6-b are provided Device, wherein：

Second acquisition unit 601, for obtaining Doppler's estimated value according to the judged result；

Second computing unit 602, for according to the signal path position, calculating maximum delay expanding value and root-mean-square time delay Expanding value；

3rd acquiring unit 603, for being expanded according to Doppler's estimated value, maximum delay expanding value and root-mean-square time delay Exhibition value, obtains current time zone channel estimation value.

The judge module 201 of 2 example of accompanying drawing can also include the 4th acquiring unit 701 and the 4th judging unit 702, such as attached The channel estimating apparatus that another embodiment of the present invention shown in Fig. 7 is provided, wherein：

4th acquiring unit 701, for carrying out equilibrium according to local training sequence and to the data in training sequence domain before Data afterwards, obtain Signal to Interference plus Noise Ratio；

4th judging unit 702, if being more than or equal to the 4th threshold value for the Signal to Interference plus Noise Ratio, judges the history Channel is close to the present channel.

The signal footpath acquisition module 202 of 7 example of accompanying drawing 2 or accompanying drawing can also include the 5th acquiring unit 801, the second para-position The channel estimation that sum unit 802 and value unit 803, such as accompanying drawing 8-a or another embodiment of the present invention shown in accompanying drawing 8-b are provided Device, wherein：

5th acquiring unit 801, for the channel estimation power P_role of different time points is obtained according to user profile （k）；

Second para-position sum unit 802, if being close to the present channel for the historical channel, by each of history Tap average channel estimation power value sequence P_mean（k）With the channel estimation power P_role of the different time points（k）Enter Row para-position summation obtains new power value sequence P ' _ mean（k）, by new power value sequence P ' and _ mean（k）Intermediate value is more than The sequence position of the Pmax/3 is taken as the signal path position, and the P ' max_mean are the new power value sequence P’_mean（k）Middle maximum, the P ' max_mean/3 for the maximum P ' max_mean 1/3；

Value unit 803, if keeping off with the present channel for the historical channel, by short at the nearest moment cycle Tap power estimates value sequence P in real time（k）Intermediate value is taken as the signal path position more than the sequence position of the Pmax/2, The Pmax is that real-time tap power of short cycle at the nearest moment estimates value sequence P（k）Middle maximum, the Pmax/2 are institute State the 1/2 of maximum Pmax.

The estimated value acquisition module 203 of 7 example of accompanying drawing 2 or accompanying drawing can also include the 6th acquiring unit 901, such as accompanying drawing 9- The channel estimating apparatus that a or another embodiment of the present invention shown in accompanying drawing 9-b are provided.6th acquiring unit 901 is for according to described Signal path position, the subscriber channel to being extracted according to user profile estimate power Che（k）The zero setting of middle non-signal path position value is obtaining Obtain the current time zone channel estimation value.

The judge module 201 of 2 example of accompanying drawing can also include correlation unit 1001 and the 5th judging unit 1002, such as accompanying drawing The channel estimating apparatus that another embodiment of the present invention shown in 10 is provided, wherein：

Correlation unit 1001, for the synchronous of history is trained code word SYNC_DL_history（k）Virtual value is same with current Step training code word SYNC_DL（k）Related operation is normalized, related operation factor beta is obtained, the related operation factor beta is

The K is channel estimation length；

5th judging unit 1002, if being more than or equal to the 5th threshold value for the related operation factor beta, judges to work as Front channel is close to historical channel, otherwise, judges that present channel is kept off with historical channel.

The signal footpath acquisition module 201 of 10 example of accompanying drawing 2 or accompanying drawing can also include the 7th acquiring unit 1101 and the 8th The channel estimating apparatus that acquiring unit 1102, such as accompanying drawing 11-a or another embodiment of the present invention shown in accompanying drawing 11-b are provided, wherein：

7th acquiring unit 1101, if for judging that the present channel is close to the historical channel, according to superposition Training sequence Steiner channel estimation powers activation footpath obtain the lower moment training sequence channel estimation signal footpath Position；

8th acquiring unit 1102, if for judging that the present channel is kept off with the historical channel, basis should The activation footpath of the Steiner channel estimation powers of the training sequence at moment obtains the letter of the channel estimation of the training sequence at lower moment Number path position.

The estimated value acquisition module 203 of accompanying drawing 2, accompanying drawing 11-a or accompanying drawing 11-b examples can also include the 9th acquiring unit 1201, such as the channel estimating apparatus that accompanying drawing 12-a, accompanying drawing 12-b or another embodiment of the present invention shown in accompanying drawing 12-c are provided.9th Acquiring unit 1201 for according to the signal path position, the value of the activation path position of Steiner channel estimations being retained, other Value zero setting is obtaining the current time zone channel estimation value.

It should be noted that the content such as information exchange, implementation procedure between each module/unit of said apparatus, due to The inventive method embodiment is based on same design, and the technique effect which brings is identical with the inventive method embodiment, particular content The narration in the inventive method embodiment is can be found in, here is omitted.

One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment is can Instruct related hardware to complete with by program, one or more or whole of such as following various methods：

Judge whether historical channel and present channel are close to；

According to the judged result of the judgement, signal path position is obtained；

According to the signal path position, current time zone channel estimation value is obtained；

It is described to judge whether historical channel and present channel are close to any knot using a kind of following mode or in the following manner Close：

By judging historical channel estimated value is judged with whether present channel estimated value is close to；Or

The parameter estimated using historical channel carries out equilibrium to current time data, is judged according to portfolio effect；Or

Historical data and current data are carried out related, judged according to related correlated results.

One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment is can Instruct related hardware to complete with by program, the program can be stored in a computer-readable recording medium, storage Medium can include：Read only memory（ROM, Read Only Memory）, random access memory（RAM, Random Access Memory）, disk or CD etc..

Above a kind of channel estimation methods provided in an embodiment of the present invention and device are described in detail, herein should The principle and embodiment of the present invention are set forth with specific case, the explanation of above example is only intended to help and manages The solution method of the present invention and its core concept；Simultaneously for one of ordinary skill in the art, according to the thought of the present invention, Will change in specific embodiment and range of application, in sum, this specification content is should not be construed as to this Bright restriction.

Claims (20)

1. a kind of channel estimation methods, it is characterised in that include：

Judge whether historical channel and present channel are close to；

Signal path position is obtained according to judged result；

According to the signal path position, current time zone channel estimation value is obtained；

It is described to judge whether historical channel and present channel are close to any combination using a kind of following mode or in the following manner：

Judged with whether present channel estimated value is close to by historical channel estimated value；Or

The parameter estimated using historical channel carries out equilibrium to current time data, is judged according to portfolio effect；Or

Historical data and current data are carried out related, judged according to related correlated results.

2. the method for claim 1, it is characterised in that described by historical channel estimated value and present channel estimated value Whether be close to and judge, including it is following any one：

If normalization coefficient α is more than or equal to the first threshold value, judge that the historical channel is close to present channel, otherwise, Judge that the historical channel is kept off with present channel, the normalization coefficient α is equal toInstitute It is channel estimation length to state N, and P (k) is that real-time tap power estimates value sequence, and P_mean (k) is averagely believed for each tap of history Estimate power value sequence in road；Or

If the absolute value of the difference of T_rms_H and T_rms is more than the exhausted of the second threshold value and the difference of T_max_H and T_max To value more than the 3rd threshold value, then judge that the historical channel is kept off with present channel, otherwise, judge the historical channel and Present channel is close to；Wherein, the T_rms_H is to estimate to ask after the zero setting of non-signal footpath in value sequence P (k) by real-time tap power The history root-mean-square delay spread value for obtaining, the T_rms are that real-time tap power being estimated, non-signal footpath is put in value sequence P (k) The root-mean-square delay spread value tried to achieve after zero, the T_max_H are that real-time tap power is estimated non-signal in value sequence P (k) The history delay estimation maximum value tried to achieve after the zero setting of footpath, the T_max are that the real-time tap power is estimated value sequence P (k) The delay estimation maximum value tried to achieve after the zero setting of middle non-signal footpath；Or

Para-position is carried out averagely to obtain current strong footpath sequence pos (s) to channel estimation power P_role (k) of different time points, If the virtual value (s) of strong footpath sequence pos_history of history is differed with the virtual value in current strong footpath sequence pos (s), Judge that the historical channel is kept off with the present channel, otherwise, judge that the historical channel is close to the present channel.

3. method as claimed in claim 1 or 2, it is characterised in that described that signal path position is obtained according to the judged result, Including：

If the historical channel is close to the present channel, by each tap average channel estimation power value sequence P_ of history With real-time tap power, mean (k) estimates that value sequence P (k) carries out para-position summation and obtains new power value sequence P ' _ mean (k)；

By new power value sequence P ' _ mean (k) intermediate values more than P ' max_mean/3 sequence position be taken as it is described Signal path position, the P ' max_mean are new power value sequence P ' maximum in _ mean (k), the P ' max_ Mean/3 for the maximum P ' max_mean 1/3；

If the historical channel is kept off with the present channel, real-time tap power is estimated into that the value in value sequence P (k) is big The signal path position is taken as in the sequence position of the Pmax/3, the Pmax is that real-time tap power estimates value sequence Maximum in P (k), the Pmax/3 are the 1/3 of the maximum Pmax.

4. method as claimed in claim 1 or 2, it is characterised in that described according to the signal path position, obtains current time zone Channel estimation value, including in the following manner any one：

The corresponding power value sequence of the signal path position is taken directly as the current time zone channel estimation value；Or,

According to the signal path position, maximum delay expanding value and root-mean-square delay spread value are calculated；

According to the maximum delay expanding value and root-mean-square delay spread value, current time zone channel estimation value is obtained；Or,

Doppler's estimated value is obtained according to the judged result；

According to the signal path position, maximum delay expanding value and root-mean-square delay spread value are calculated；

According to Doppler's estimated value, maximum delay expanding value and root-mean-square delay spread value, obtain current time zone channel and estimate Evaluation.

5. the method for claim 1, it is characterised in that the parameter that the employing historical channel is estimated is to current time number According to equilibrium is carried out, judged according to portfolio effect, including：

The data after equilibrium are carried out according to local training sequence and to the data in training sequence domain before, Signal to Interference plus Noise Ratio is obtained；

If the Signal to Interference plus Noise Ratio is more than or equal to the 4th threshold value, judge that the historical channel is close to the present channel.

6. the method as described in claim 1 or 5, it is characterised in that described that signal path position is obtained according to the judged result, Including：

Channel estimation power P_role (k) of different time points is obtained according to user profile；If the historical channel is worked as with described Front channel is close to, then by each tap average channel estimation power value sequence P_mean (k) of history and the different time points Channel estimation power P_role (k) carries out para-position summation and obtains new power value sequence P ' _ mean (k), by the new power Value sequence P ' _ mean (k) intermediate values are taken as the signal path position, the P ' max_ more than the sequence position of the Pmax/3 Mean is new power value sequence P ' maximum in _ mean (k), the P ' max_mean/3 are the maximum P ' max_ The 1/3 of mean；

If the historical channel is kept off with the present channel, by real-time in short at nearest moment cycle tap power estimated value sequence Row P (k) intermediate value is taken as the signal path position more than the sequence position of the Pmax/2, when the Pmax is described nearest Carve real-time tap power of short cycle and estimate maximum in value sequence P (k), the Pmax/2 is the 1/2 of the maximum Pmax.

7. the method as described in claim 1 or 5, it is characterised in that described according to the signal path position, obtains current time zone Channel estimation value includes：

According to the signal path position, the subscriber channel to being extracted according to user profile estimates non-signal footpath position in power Che (k) Put value zero setting to obtain the current time zone channel estimation value.

8. the method for claim 1, it is characterised in that it is described historical data and current data are carried out it is related, according to Correlated results judgement, including：

Synchronous training code word SYNC_DL_history (k) virtual value of history is trained into code word with current sync SYNC_DL (k) is normalized related operation, obtains related operation factor beta, and the related operation factor beta isThe K is channel estimation length；

If the related operation factor beta is more than or equal to the 5th threshold value, judge that present channel is close to historical channel, it is no Then, judge that present channel is kept off with historical channel.

9. the method as described in claim 1 or 8, it is characterised in that described that signal path position is obtained according to the judged result, Including：

If judging, the present channel is close to the historical channel, is estimated according to the Steiner channels of the training sequence of superposition The activation footpath of meter power obtains the signal path position of the channel estimation of the training sequence at lower moment；

If judging, the present channel is kept off with the historical channel, according to the Steiner letters of the training sequence at the moment Estimate that the activation footpath of power obtains the signal path position of the channel estimation of the training sequence at lower moment in road.

10. the method as described in claim 1 or 8, it is characterised in that described according to the signal path position, when obtaining current Domain channel estimation value, including：

According to the signal path position, the value of the activation path position of Steiner channel estimations is retained, other values zero setting is obtaining The current time zone channel estimation value.

11. a kind of channel estimating apparatus, it is characterised in that described device includes：

Judge module, for judging whether historical channel and present channel are close to；

Signal footpath acquisition module, for obtaining signal path position according to the judged result；

Estimated value acquisition module, for according to the signal path position, obtaining current time zone channel estimation value；

Using any combination of a kind of following mode or following manner, the judge module judges that historical channel and present channel are It is no to be close to：

Judged with whether present channel estimated value is close to by historical channel estimated value；Or

The parameter estimated using historical channel carries out equilibrium to current time data, is judged according to portfolio effect；Or

Historical data and current data are carried out it is related, according to correlated results judge.

12. devices as claimed in claim 11, it is characterised in that when the judge module is used for by historical channel estimated value Whether it is close to present channel estimated value when judging, the judge module includes：

First judging unit, if being more than or equal to the first threshold value for normalization coefficient α, judges the historical channel and works as Front channel is close to, and otherwise, judges that the historical channel is kept off with present channel, and the normalization coefficient α is equal toThe N is channel estimation length, and P (k) is that real-time tap power estimates value sequence, P_ Each tap average channel estimation power value sequences of the mean (k) for history；Or

Second judging unit, if being more than second threshold value and T_max_H with the absolute value of the difference of T_rms for T_rms_H It is more than the 3rd threshold value with the absolute value of the difference of T_max, then judges that the historical channel is kept off with present channel, otherwise, Judge that the historical channel is close to present channel, the T_rms_H is that real-time tap power is estimated non-in value sequence P (k) The history root-mean-square delay spread value tried to achieve after the zero setting of signal footpath, the T_rms are that real-time tap power is estimated value sequence P K the root-mean-square delay spread value tried to achieve after the zero setting of non-signal footpath in (), the T_max_H are by real-time tap power estimated value sequence The history delay estimation maximum value tried to achieve after the zero setting of non-signal footpath in row P (k), the T_max is by the real-time tap power The delay estimation maximum value tried to achieve after the zero setting of non-signal footpath in estimation value sequence P (k)；Or

3rd judging unit, carries out para-position for channel estimation power P_role (k) to different time points and is averagely worked as with obtaining Front strong footpath sequence pos (s), if the virtual value (s) of strong footpath sequence pos_history of history and current strong footpath sequence pos (s) Virtual value differ, then judge that the historical channel is kept off with the present channel, otherwise, judge the historical channel with The present channel is close to.

13. devices as described in claim 11 or 12, it is characterised in that signal footpath acquisition module includes：

First para-position sum unit, if being close to the present channel for the historical channel, each tap of history is put down With real-time tap power, channel estimation power value sequence P_mean (k) estimates that value sequence P (k) carries out para-position summation and obtains new Power value sequence P ' _ mean (k), if the historical channel is kept off with the present channel, real-time tap power is estimated Value in value sequence P (k) is taken as the signal path position more than the sequence position of the Pmax/3, and the Pmax is real-time Tap power estimates maximum in value sequence P (k), and the Pmax/3 is the 1/3 of the maximum Pmax, by the new power Value sequence P ' _ mean (k) intermediate values are taken as the signal path position, the P ' more than the sequence position of P ' max_mean/3 Max_mean is new power value sequence P ' maximum in _ mean (k), the P ' max_mean/3 are the maximum P ' The 1/3 of max_mean.

14. devices as described in claim 11 or 12, it is characterised in that the estimated value acquisition module is directly taking the letter Number corresponding power value sequence of path position is used as the current time zone channel estimation value；Or, the estimated value acquisition module bag Include：

First computing unit, for according to the signal path position, calculating maximum delay expanding value and root-mean-square delay spread value；

First acquisition unit, for according to the maximum delay expanding value and root-mean-square delay spread value, obtaining current time zone letter Road estimated value；Or,

The estimated value acquisition module includes：

Second acquisition unit, for obtaining Doppler's estimated value according to the judged result；

Second computing unit, for according to the signal path position, calculating maximum delay expanding value and root-mean-square delay spread value；

3rd acquiring unit, for according to Doppler's estimated value, maximum delay expanding value and root-mean-square delay spread value, obtaining Take current time zone channel estimation value.

15. devices as claimed in claim 11, it is characterised in that when the judge module is used for what is estimated using historical channel Parameter carries out equilibrium to current time data, and when being judged according to portfolio effect, the judge module includes：

4th acquiring unit, for carrying out the number after equilibrium according to local training sequence and to the data in training sequence domain before According to acquisition Signal to Interference plus Noise Ratio；

4th judging unit, if for the Signal to Interference plus Noise Ratio be more than or equal to the 4th threshold value, judge the historical channel with The present channel is close to.

16. devices as described in claim 11 or 15, it is characterised in that signal footpath acquisition module includes：

5th acquiring unit, for channel estimation power P_role (k) of different time points is obtained according to user profile；

Second para-position sum unit, if being close to the present channel for the historical channel, each tap of history is put down Channel estimation power value sequence P_mean (k) carries out para-position with channel estimation power P_role (k) of the different time points _ mean (k) intermediate values that summation obtains new power value sequence P ' _ mean (k), by new power value sequence P ' are more than described The sequence position of Pmax/3 is taken as the signal path position, and the P ' max_mean are new power value sequence P ' _ Maximum in mean (k), the P ' max_mean/3 for the maximum P ' max_mean 1/3；

Value unit, if being kept off with the present channel for the historical channel, short at the nearest moment cycle is taken out in real time Head power estimation value sequence P (k) intermediate value is taken as the signal path position more than the sequence position of the Pmax/2, described Pmax estimates maximum in value sequence P (k) for real-time tap power of short cycle at nearest moment, the Pmax/2 for it is described most The 1/2 of big value Pmax.

17. devices as described in claim 11 or 15, it is characterised in that the estimated value acquisition module includes：

6th acquiring unit, for according to the signal path position, the subscriber channel to being extracted according to user profile estimates power In Che (k), the zero setting of non-signal path position value is obtaining the current time zone channel estimation value.

18. devices as claimed in claim 11, it is characterised in that when the judge module is for by historical data and current number According to correlation is carried out, when being judged according to correlated results, the judge module includes：

Correlation unit, for the synchronous of history is trained code word SYNC_DL_history (k) virtual value and current sync Training code word SYNC_DL (k) is normalized related operation, obtains related operation factor beta, and the related operation factor beta isThe K is channel estimation length；

5th judging unit, if for the related operation factor beta be more than or equal to the 5th threshold value, judge present channel and Historical channel is close to, and otherwise, judges that present channel is kept off with historical channel.

19. devices as described in claim 11 or 18, it is characterised in that signal footpath acquisition module includes：

7th acquiring unit, if for judging that the present channel is close to the historical channel, according to the training sequence of superposition The activation footpath of the Steiner channel estimation powers of row obtains the signal path position of the channel estimation of the training sequence at lower moment；

8th acquiring unit, if for judging that the present channel is kept off with the historical channel, according to the instruction at the moment The activation footpath of the Steiner channel estimation powers of white silk sequence obtains the signal footpath position of the channel estimation of the training sequence at lower moment Put.

20. devices as described in claim 11 or 19, it is characterised in that the estimated value acquisition module includes：

9th acquiring unit, for according to the signal path position, the value of the activation path position of Steiner channel estimations being protected Stay, other values zero setting is obtaining the current time zone channel estimation value.